PGS and the guru pyramid

Two things, not one

The word "pyramid" gets used for both of these, and they are different measurements.

PGS — the Pythia Guru Score is a 0-100 number. It runs 14 hurdles drawn from named investing traditions over a company's figures and reports the percentage of evaluable hurdles it passes. Hurdles whose inputs are missing are excluded from the denominator rather than counted as failures, and a company with fewer than 8 evaluable hurdles gets no score at all instead of a misleading one built on four data points.

The pyramid is a different view over an 18-signal engine, and its output is a pass count from 0 to 18 plus a tier. It exists to sort the whole universe into a shape you can look at, which a 0-100 score cannot do.

If a page shows you "PGS 61", that is the percentage figure. If it shows you "12 of 18", that is the pyramid.

The four tiers

The pyramid places a company by how many of the 18 signals fire:

Worth knowing before you go looking for a perfect score: nobody in the universe passes more than 16 of the 18, and only a handful reach that. The signals include several that are mutually awkward — a company cheap enough to clear the deep-value hurdles is rarely also growing fast enough to clear the growth ones. That tension is deliberate. A screen everything passes is not a screen.

What the hurdles actually test

The 14 PGS hurdles are concrete rather than thematic. They include operating and financial leverage, debt against EBITDA and against net income, EBIT per employee, the dollar-retained-to-market-value test, owner-earnings growth, capital spending against earnings, return on capital, earnings yield, price against the Graham Number, the PEG ratio, and price to sales.

Every one of them traces to a published framework, which is the point — the score is arguable because each component is.

How it differs from PAS

PAS asks an empirical question: relative to companies like this one, how do the fundamentals rank? It is sector-neutral by construction and says nothing about absolute standards.

PGS asks a normative one: does this clear the bars that named investors actually set? Those bars are absolute. A debt-to-EBITDA hurdle does not care that a company's peers are all leveraged too.

The two disagree often, and the disagreement is informative rather than a defect. A company can score PAS 78 and pass only 5 of 18 pyramid signals — academically strong against its peers while failing most absolute value tests, which is the normal state of a good business at a high price. The reverse happens too: pyramid 14 with PAS 42 is the classic deep-value setup, cheap enough to clear absolute bars while ranking poorly against peers on quality.

What neither of them is

Neither is a recommendation. Both are summaries of published tests applied to reported figures, and every component opens into the inputs behind it. The useful question is never "what is the score" but "which test is it failing, and do I agree that the test matters here".